Efficient estimation of die-level process parameter variations via the EM-algorithm

A new approach for efficient estimation of die-level process parameter variations based on the expectation- maximization algorithm is proposed. To estimate the parameters and enhance diagnostic analysis, dedicated embedded sensors have been designed. Additionally, to guide the test with the information obtained through monitoring process variations, maximum-likelihood method and adjusted support vector machine classifier is employed. The information acquired is re-used and supplement the circuit calibration. The proposed estimation method is evaluated on a prototype ADC converter with dedicated sensors fabricated in standard single poly, five metal 0.09-mum CMOS